It is proposed to obtain realistic and credible risk estimates for breast-cancer mortality due to clinical mammographic imaging examinations. Given the increasing emphasis on clinical mammographic screening for breast cancer, it is of societal importance to provide realistic risk estimates - with realistic confidence bounds - for breast cancer induction from routine mammographic X rays. Concern about this risk is a major factor, both from the perspective of the patient and the family physician, in non compliance with suggested guidelines for routine screening. Previous estimates of this risk have a) not accounted for the significantly increased risk from the very low-energy X rays used in screening mammography and b) based their estimates on only a single projected breast-cancer risk model. Direct studies of mammographic screening programs have insufficient power to quantitate risks of mammographic screening, so estimates of the risks must be based on extrapolations from other cohorts (A-bomb survivors, medical irradiation). These other cohorts, however, were exposed to far higher- energy X or gamma rays than the very low-energy X rays used in mammograms; dose-for-dose, low-energy X rays are considerably more radiobiologically damaging than high-energy X or gamma rays. So it is likely that this risk estimation process underestimates risks of screening mammograms - preliminary estimates are by a factor of 1.5-2. State-of-the art radiobiological techniques will be used to extrapolate from those situations where radiation-induced breast-cancer risk estimates are better known (A-bomb survivors, medical irradiations), to risks of a screening mammograms. Experimental microdosimetric measurements will be used to characterize the energy-deposition characteristics of various mammographic X-ray beams in current or projected use. Based on these data, the biological effectiveness of these mammographic beams will be estimated relative to those for which breast-cancer risks have been estimated (e.g. A-bomb survivors, TB patients). The bottom line of these studies will be realistic estimates of the risk from routine screening mammograms, together with a sensitivity analysis of the confidence limits that may reasonably be associated with these risk estimates.